4.3 Neuroprotection by Trichostatin A (TSA) pre-treatment
4.3.4 Enhancement of histone acetylation by Trichostatin A renders neurons more
The following experiment was performed to investigate whether enhancement of histone H4 acetylation by TSA is causally linked to its neuroprotective effects. Rat primary cortical
Histone acetylation and neuroprotection Results
neurons were pre-treated with 300nM TSA for 24 hours and then subjected to 120min OGD.
Immunoctytochemical staining was performed after 24 hours, with antibodies for the neuronal marker microtubule associated protein MAP2 (green), acetylated (Ac) histone H4 (red) and nuclear Hoechst 33258 dye (blue). As presented in the photomicrographs of Figure 17B different levels of histone H4 acetylation was observed in individual neurons. Neurons were counted and classified in groups according to the acetylation levels of histone H4 in their nuclei, low- vs. high-acetylated, and the result was presented as the figure 17A. Naive neuronal cultures displayed less than 20% high-acetylated neurons, upon TSA treatment this number was dramatically increased to more than 80%. MAP2 is a neuronal marker which has been previously reported to be very sensitive to ischemic insult (Harms et al. 2001).
Following OGD the number of MAP2-positive, healthy, neurons was three-fold higher in TSA pre-treated cultures, demonstrating once again TSA’s neuroprotective properties.
Nevertheless, the most important finding of this experiment was that neurons surviving the ischemic insult both in vehicle- and TSA pre-treated cultures had high histone H4 acetylation levels in their nuclei. Noteworthy, dead neurons which show morphology characteristic to apoptosis also presented an intense signal after acetylated histone H4 staining. Taken all together, these data however suggest that high-acetylated neurons are more likely to survive an ischemic insult.
A
Histone acetylation and neuroprotection Results
B
Figure 17 Trichostatin A enhances histone acetylation levels in cortical neurons. (A, B) Primary cortical neurons were pre-treated with 300 nM TSA for 24 hours and then subjected to 150 min of oxygen-glucose deprivation (OGD). At 24 hours, immunocytochemical staining was performed with antibodies raised against the neuronal marker microtubule-associated protein Map-2 (green), acetylated (Ac)-histone H4 (red) and Hoechst 33258 (blue). The fractions of dead Map-2-positive neurons (dagger), low-acetylated living Map-2-positive neurons (open arrow) and high-acetylated living Map-2-positive neurons (filled arrow) were determined.
Twohundred Map-2-positive neurons per group were evaluated and divided into three different groups (high-acetylated living neurons, low-(high-acetylated living neurons and dead neurons) using a fluorescence microscope and a digital camera without changing the lightexposure time during the process. The acetylation status was evaluated in Map-2-positive viable neurons with dendrites and intact nuclear morphology (visualized by Hoechst DNA counterstaining), whereas dead neurons and those exhibiting nuclear shrinkage, blebbing or chromatin condensation were not considered. Counts were performed in two independent experiments with at least three different cover slips per group. Scale bar, 30 µm.
Histone acetylation and neuroprotection Results
4.3.5 Trichostatin A up-regulates gelsolin expression in cortical neurons and in mice brain
Our group has previuosly reported anti-apoptotic and anti-excitotoxic properties for gelsolin protein (Harms et al. 2004). Previously, my MSc thesis presented that TSA enhanced histone H4 acetylation levels at gelsolin gene promoter region in primary cortical cultures and that the subsequent result of this event was increase in gelsolin mRNA levels (Meisel et al. 2006).
Here, to test whether TSA’s effect on gelsolin transciption is also translated into its protein levels, rat primary cortical cultures were treated with different concentrations of TSA for 24 hours and with 300nM TSA for different durations. Cellular protein extracts were subjected to SDS-PAGE, followed by western immunoblotting using antibodies against murine gelsolin and histone H4. TSA up-regulated gelsolin protein levels in a time and concentration dependent manner (Figure 18A). Importantly, these increases were in agreement with the enhancement of histone H4 as well as histone H3 acetylation by TSA treatment (Figure 16). A significant up-regulation of gelsolin was observed with TSA only at the doses of 300nM and 500nM and with the treatment durations for 12 and 24 hours. TSA treatment with lower doses, 25nM-100nM, as well as for shorter time duration, 6 hours, did not result in any prominent up-regulation of gelsolin protein, neither did they afford neuroprotection against OGD. Anti-histone H4 antibody was utilized for the demonstration of equal protein loading.
To address the question of whether up-regulation of gelsolin protein and enhancement of histone acetylation levels occur within the same individual neurons, immunocytochemistry was performed on cortical cultures treated with 300nM TSA for 24 hours and antibodies against gelsolin (green), acetylated (Ac) histone H4 (red) and Hoechst 33258 dye (blue) were used. Immunocytochemistry followed by high resolution confocal laser scanning microscopy demonstrated that individual neurons with high histone H4 acetylation pattern in their nuclei also expressed increased gelsolin levels in their cell body (Figure 18B).
To examine whether TSA exerts a similar effect on gelsolin expression also in mice brain, 129/Sv mice were treated with TSA at the doses of 1 and 5mg/kg body weight for fourteen days, by daily intraperitoneal injections. Thereafter, whole brain cellular extracts were used for western immunoblotting. Figure 18C shows dose dependent up-regulation of gelsolin protein in mice brain upon TSA treatment. Importantly 1mg/kg TSA, a dose which neither
Histone acetylation and neuroprotection Results
afforded neuroprotection nor enhanced histone H4 acetylation in vivo, did not up-regulate gelsolin protein in mice brain. Anti-actin antibody was utilized for the demonstration of equal protein loading. Immunohistochemistry was carried out on brain slices of mice confirmed the up-regulation of gelsolin after the neuroprotective TSA treatment and suggested a neuron specific expression pattern (Figure 18D). Double-labeling fluorescent immunohistochemistry indeed demonstrated the up-regulation of gelsolin in neuronal marker (NeuN+) positive neurons (Figure 18E). The photomicrographs were taken from a neuron-rich brain region, hippocampus, to better present up-regulation of gelsolin protein by TSA.
A
B
Histone acetylation and neuroprotection Results
C D
E
Figure 18 Trichostatin A up-regulates gelsolin expression in cortical cultures and in mice brain. Primary cortical neurons were treated with different concentrations of TSA for the indicated durations and were subjected to western immunoblotting. (A) Twenty µ g of protein was subjected to sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and incubated with polyclonal antibodies against murine gelsolin and histone H4 after stripping and reprobing the same membrane. (B) Primary cortical neurons were seeded on glass cover slips and treated with 300 nM TSA on the ninth day in vitro (DIV9). Cells were fixed after 24 hours and immunocytochemistry was performed. Gelsolin (green), ac-histone H4 (red) and Hoechst staining (blue) were
Histone acetylation and neuroprotection Results
immunohistochemical stainings of sections from frontoparietal cortex and hippocampus showing the effect of TSA treatment on gelsolin expression in brains of 129/Sv mice treated with TSA at 1 and 5 mg/kg body weight daily given intraperitoneally for 14 days. (E) Up-regulation of gelsolin (red) by TSA in neurons of hippocampal CA1 region. Neuronal marker NeuN (green). Scale bars, (D) 100 µm (cortex) and 25 µm (hippocampus), (E) 20 µm. Representative results of three independent experiments with similar results are shown.
4.3.6 Trichostatin A confers actin filament remodelling
Gelsolin was previuosly reported to sever actin microfilaments, conferring remodelling of actin cyctoskeleton (Kwiatkowski et al. 1988, Harms et al. 2004). We therefore examined whether filamenteous actin dynamics are altered upon TSA-induced gelsolin up-regulation.
Rat cortical cultures were treated with 300nM TSA for different durations, followed by phalloidin-fluorescein staining. The photomicrographs indeed revealed that filamenteous actin levels were reduced in parallel to enhancement of histone H4 acetylation after TSA treatment (Figure 19A). Importantly, these two events occured within the same individual neurons as the merged photomicrographs clearly demonstrate. Methanol extraction of bound phalloidin and subsequent photometric quantification, using a fluorescence plate reader, revealed significant decrease in filamenteous actin levels by 300nM TSA treatment (Figure 19B).
Noteworthy is that the decrease in filamentous actin levels reached statistical significance only after TSA treatments for 12 hours or longer.
We further tested whether TSA’s effect on actin cytoskeleton still persists in neurons after their exposure to oxygen glucose deprivation (OGD). For this purpose, rat primary cortical cultures were pre-treated with 300nM TSA for 12 hours, subjected to OGD and subsequently cells were fixed at different time intervals. Phalloidin-fluorescein staining of actin cytoskeleton was followed by methanol extraction of F-actin bound phalloidin. Data obtained from photometric quantification is presented as figure 19C. OGD decreased actin microfilament levels to approximately 80% of baseline, yet TSA-induced differences in filamentous actin levels were sustained at all time points following OGD.
Histone acetylation and neuroprotection Results
A B
C
Figure 19 Effects of Trichostatin A on filamentous actin. Primary cortical neurons were seeded on glass cover slips and pre-treated with 300 nM TSA or vehicle (control) for 12 hours (A, C) or 6, 12 and 24 hours (B). (A) Following end of the treatment, neurons were fixed with 4% paraformaldehyde, stained with phalloidin–
fluorescein (green) and immunostained against acetylated (Ac)-histone H4 (red). Scale bar, 30 µm. (B) Phalloidin–fluorescein was extracted and fluorescence was quantified. *P < 0.001 vs. vehicle treatment. (C) In a different subset of experiments, at 12 hours following TSA or vehicle (control) treatment, cultured neurons were subjected to oxygen-glucose deprivation (OGD) or normoxia as control treatment, and subsequently fixed with 4% paraformaldehyde at the indicated time points, and stained with phalloidin–fluorescein. Phalloidin–
fluorescein was then extracted by methanol and fluorescence signal was quantified. *P < 0.0001 vs. vehicle
Histone acetylation and neuroprotection Results
4.3.7 Trichostatin A reduces intracellular calcium overload caused by ischemic injury
Intracellular calcium overload is considered to be a central player in ischemic neuronal cell death (Choi, 1995). Our group has previously reported that gelsolin exerts anti-excitotoxic effects following cerebral ischemia via remodelling actin cytoskeleton and also stabilising calcium channels (Endres et al. 1999). Here, whether TSA has effect on intracellular calcium overload triggered by oxygen-glucose deprivation was investigated. Rat primary cortical cultures were pre-treated with 300nM TSA for 12 hours and subjected to OGD on in vitro day 9 (DIV9). It is well known that major calcium influx take place both during the ischemic event and also directly after reperfusion/re-oxygenation. Thus the cultures were loaded with calcium binding fluorescent dye Fluo-4 AM either during OGD or immediately after OGD
and subsequently measured the fluorescent signal after a dye incubation period of 45min. In either paradigm significantly lower intracellular calcium levels were observed when the cultures received TSA treatment prior to OGD (figure 20A and 20B). Moreover, the differences in fluorescence signal intensities between TSA and vehicle pre-treated cultures persisted throughout subsequent repetitive measurements at 10min intervals for one hour after the first measurement (data not shown).
A B
Histone acetylation and neuroprotection Results
Figure 20 Trichostatin A decreases calcium influx following combined oxygen-glucose deprivation.
Primary cortical neurons were pre-treated with 300 nM TSA for 24 hours and subjected to 150 min OGD.
Calcium indicator fluorescent dye Fluo-4 AM was loaded either during (A) or immediately after (B) OGD.
Fluorescent signal intensities were measured following an incubation period of 45 min. Values (mean±SEM) in relative fluorescence units (RFU) are 90.4±3.7, 92.4±3.9, 550.9±1.3, and 358.1±1.6 in (A), and 133.9±2.3, 136.8±2.5, 453.6±1.2, and 334.2±1.1 in (B) *P < 0.001 for TSA vs. vehicle.
4.3.8 Trichostatin A prevents loss of mitochondrial membrane potential caused by ischemic injury
After assessment of intracellular calcium levels, we measured loss of mitochondrial membrane potential by tetramethyl rhodamine ethyl ester (TMRE) fluorescence, which is an event implicated in cell death cascades mediated by the mitochondria. For this purpose, rat primary cortical cultures were pre-treated with 300nM TSA for 12 hours and subjected the cultures to oxygen-glucose deprivation on in vitro day 9 (DIV9). Twenty-four hours afterwards TMRE was added into culture medium and photomicrographs were taken following an incubation period of 40min. Figure 21A demonstrates the loss of mitochondrial membrane potential, by means of loss of fluorescence signal intensity, however this loss was clearly prevented in cultures which were treated with TSA prior to OGD. Subsequent measurements of fluorescent signals are presented as figure 21B. Four different OGD durations carried out on sister cultures yielded the result that only longer durations of OGD caused statistically significant loss of mitochondrial membrane potential, although the loss was detectable after all four OGD durations. Importantly, OGD-caused mitochondrial dysfunction was prevented when cultures were pre-treated with TSA. TSA‘s effect was evident in all durations of OGD, milder as well as more severe forms of the ischemic insult.
Neuroprotection by TSA was also confirmed in experiments by measurement of lactate dehydrogenase (LDH) release into the medium (data not shown).
Histone acetylation and neuroprotection Results
A
B
Figure 21 Trichostatin A decreases loss of mitochondrial membrane potential following combined oxygen/glucose deprivation. Primary cortical neurons were pre-treated with 300 nM TSA for 12 hours and
Histone acetylation and neuroprotection Results
subjected to various durations of OGD. (A) Loss of mitochondrial membrane potential was measured 24 hours after OGD. TMRE was added into the culture wells and photomicrographs were taken after an incubation period of 40 min. (B) Subsequently fluorescent signal was measured in a multiwell fluorescence plate reader. There were no significant differences between TSA vs. vehicle pre-treated cultures in the control conditions. n=8.
*P<0.001 for TSA vs. vehicle in each OGD condition. #P<0.05 for corresponding vehicles in control vs. OGD 2:30 hours and control vs. OGD 2:45 hours. Scale bar, 100µm.
4.3.9 Trichostatin A pre-treatment does not protect gelsolin-deficient mice against brain ischemic injury
My MSc thesis has previously presented TSA’s failure in protecting gelsolin-deficient murine neuronal cultures against ischemic insult, underscoring gelsolin as an integral perpetrator of TSA’s neuroprotective effects (Meisel et al., 2006). Here, we tested whether gelsolin has a similar role in neuroprotection by TSA in vivo. For this purpose, gelsolin-deficient and wild-type mice were pre-treated with the neuroprotective TSA pre-treatment regime, 5mg/kg body weight of TSA for fourteen days by daily intraperitoneal injections. Subsequently mice underwent operation of filamentous middle cerebral artery occlusion for 1 hour. After a reperfusion period of 24 hours, animals were sacrificed and brains were snap-frozen. Cryostat sectioning of the brains and hematoxylin staining of the sections were followed by determination of cerebral lesion volumes by computer-assisted volumetry on serial coronal brain sections.
TSA pre-treatment did not afford neuroprotection of gelsolin-deficient mice against MCAo for 1 hour. TSA as well as vehicle pre-treated mice had similar cerebral lesion volumes and anterior to posterior lesion areas (Figure 22A and B). An indirect method for measuring lesion volume, corrected for brain oedema, did not change the lack of protection by TSA in gelsolin-deficient mice (60+/-14 vs 58+/-8 mm³ in TSA vs. vehicle pre-treated mice, n=10 per group).
Histone acetylation and neuroprotection Results
A B
Figure 22 Trichostatin A does not protect gelsolin knockout mice against MCAo/reperfusion. The effect of Trichostatin A (TSA) pre-treatment (5 mg/kg body weight daily for 14 days given intraperitoneally) on cerebral infarct volume (A) and area (B) following 1 hour filamentous MCA occlusion and 24 hours of reperfusion compared with vehicle-injected gelsolin knockout mice. Cerebral infarction volume was determined quantitatively on 20 µm-thick, hematoxylin stained brain cryostat sections, as described previously (Endres et al., 1999). n=10 animals per groups.
4.3.10 TSA confers actin remodelling in wild-type but not in gelsolin-deficient mice brain
Gelsolin severs actin microfilaments, conferring remodelling of actin cyctoskeleton (Kwiatkowski et al. 1988, Harms et al. 2004). As presented above, Trichostatin A up-regulation of gelsolin protein was followed by significantly reduced levels of filamentous actin in rat primary cortical neurons (figure 19). Here, we examined whether TSA’s modulatory effect on actin cytoskeleton is present also in brains of wild-type vs. gelsolin-deficient mice.
Wild-type and gelsolin-deficient mice were treated with 5mg/kg body weight of TSA, or vehicle, for fourteen days by daily intraperitoneal injections. After last injections, animals were sacrificed and brains were snap-frozen. Phalloidin-fluorescein as well as nuclear Hoechst 33258 staining was carried out on coronal brain sections. Representative photomicrographs which were chosen from a neuron-rich brain region, hippocampal CA1
Histone acetylation and neuroprotection Results
area, are presented as figure 23. In wild-type mice TSA reduced filamentous actin levels, by means of reduction in fluorescent signal intensity, confirming our in vitro results. As expected, vehicle treated gelsolin-deficient mice appeared to have higher levels of filamentous actin in comparison to vehicle treated wild-type mice. Unlike in wild-type mice, however, TSA treatment did not reduce filamentous actin levels in gelsolin-deficient mice.
Figure 23 Effects of Trichostatin A on filamentous actin levels in mice brain tissue.
Phalloidin stainings (green) on 20 µm-thick coronal brain sections from wild-type and gelsolin-deficient mice treated with TSA 5mg/kg body weight for 14 days. The images show neurons from hippocampal CA1 region, counterstained with Hoechst 33258 nuclear dye (blue). Scale bar, 10 µm.